PSI - Issue 62

D. Ganora et al. / Procedia Structural Integrity 62 (2024) 653–660 Author name / Structural Integrity Procedia 00 (2019) 000–000

655

3

• Flood hazard maps available: o No levee: joint use of flood maps + DTM o Levee: consider levee elevation + 20 cm • No flood hazard maps available:

o River with possible lateral expansion: consider bank elevation + 50 cm o River with no possible lateral expansion: evaluate max discharge + hydraulic model

Table 1. Method to evaluate flood water elevation: example for the Piemonte Region (Italy). Type of data Availability

Applicability to estimate water elev.

Maps of accurate water depth/elevation estimates

Currently only at some hot spots (main cities and high-risk areas)

Applicable as the water level is reported at the section of interest

Water elevation at different sections along the river

Available for major rivers

Depending on the proximity of the bridge to the section; bridge deck elevation may require topographic survey and not only quick evaluation Maps can be compared with high resolution DTM and reliability of water elevation estimate strongly depends on the quality of topographic data

Use of hazard inundation maps

Major rivers and some minor rivers

The latter case is the most critical as it requires to evaluate: i) the flood discharge for a return period of 200 years and ii) to compute the water elevation corresponding to the design discharge. Both the steps are expected to be done in a simplified way. The guidelines provide non-compulsory suggestions: i) the discharge can be evaluated with “commonly used” formulas from the literature and ii) water level can be estimated as the normal depth of the flow. Focusing on point i), many empirical equations are available in hydrological handbooks, but with several disadvantages: first, they are usually derived as envelope curves or recorded floods, thus they are not probability based; second, they are not up-to-date (some have been developed in the 1930’s); finally, they are “valid” over very large areas and are not able to account for different catchment characteristics (area, elevation, land use, …). Moreover, such equations lead to very variable results, depending on the chosen parameterization. Another option is to use regional models, which are statistically-based relations allowing to estimate the design flood discharge in ungauged basins (e.g., Laio et al, 2011), as for example the models included in the Italian “Va.Pi” initiative. This class of model is much more reliable than empirical equations, but usually the application in small and very small catchments is not recommended because they are calibrated mainly on large ones. Moreover, their use may require some non-standard GIS applications. A further option and largely used approach for small catchments is the well-known rational formula (e.g., Maidment, 1992), that reads: ( ) = ⋅ ⋅ ( ! , ) ⋅ (1) where is the runoff coefficient that accounts for the permeability of the soil, is the areal reduction factor that accounts for the extent of the precipitation event with respect to catchment size, is the average intensity of precipitation for to the critical duration ! , given the return period , and is the catchment area. The rational formula, widely used in the engineering practice, has been criticized due to the subjectivity of its parameters (in particular and ! , see e.g. Grimaldi et al., 2012) and for the strong assumption that the return period of discharge and precipitation are assumed equal. Despite this, it is a valid tool for this kind of fast assessment thanks to its ease of use that make